EMPIRICAL ESTIMATION OF GROUNDWATER RECHARGE IN … · Olubunmi Adegun Department of Geography, University of Lagos, Lagos, Nigeria Correspondence:[email protected] ... Nigeria (Edmunds,

Journal of Geographic Issues, Vol. 2 No. 1, 2014

58

EMPIRICAL ESTIMATION OF GROUNDWATER RECHARGE IN PARTS OF

THE BASEMENT COMPLEX OF NORTHERN NIGERIA

Olubunmi Adegun

Department of Geography, University of Lagos, Lagos, Nigeria

Correspondence:[email protected]

Abstract Knowledge of recharge estimates is essential to the sustainable development and

management of groundwater resources. In this study two empirical methods were employed

in estimating of groundwater recharge in Gusau and Bauchi in semi-arid Northern Nigeria.

The recharge estimates were based on the methods developed by Sinha and Sharma (1988)

and Turc (1954).Variability of annual rainfall and recharge estimates were determined using

the coefficient of variation, while the equality of variance of recharge estimates obtained

from the empirical methods were determined using Levnne’s Test. Simple Linear Regression

was used to predict the mean groundwater recharge for the study period. Results of

groundwater recharge for Bauchi based on the method developed by Sinha and Sharma gave

a maximum and minimum estimate of 262.64mm and 158.50mm, while the highest and

lowest estimate for Gusau was 252.49mm and 142.29mm. Highest and lowest annual average

recharge estimate based on Turc’s method was, 687.49mm and 170.13mm for Bauchi, while

the maximum and minimum estimate for Gusau was 578.76mm and 69.52mm. The results of

the coefficient of variation shows a high level of variability based on Turc’s method, and a

moderate degree of variability based on Sinha and Sharma’s method. The results of Levene’s

Test indicated that variance of recharge estimates obtained from both methods differ

significantly from each other. Regression model for the two locations revealed that rainfall

generally accounts for more than 90 percent of groundwater recharge variation. The study

recommended that scientific and periodic assessments of groundwater resources potential

should be embarked upon.

Keywords: Recharge; Groundwater; Semi-Arid; Basement Complex; Rainfall.

INTRODUCTION Groundwater recharge is a process by which water enters into the phreatic or saturated zone.

It is the downward flow or percolation of water to the water table, adding to water in the

groundwater reservoir (Sharp, 2007; de Vries and Simmers, 2002). The highly variable

nature of groundwater recharge and its pivotal role in the sustainable development and

management of groundwater resources therefore makes the quantification of natural recharge

very important. In arid and semi-arid regions of the world where rates of natural groundwater

recharge are generally low in comparison with average annual rainfall, quantification of the

rate of natural groundwater recharge is necessary for the sustainable use and management of

groundwater.

In the semi-arid areas of Northern Nigeria where groundwater is being increasingly utilized,

the estimation of groundwater available for use is very important in order to prevent some of

the negative environment consequences associated with over-abstraction and mining of

groundwater. The increased reliance on groundwater in many parts of Northern Nigeria is

attributed to a number of reasons. These include the highly seasonal nature of rainfall, low

rainfall amounts and water balance deficit, increasing population, inadequate pipe-borne

water supply for domestic and municipal uses, and over- abstraction of groundwater for


59

agriculture and other uses. Furthermore, the downward trend of river discharge occasioned by

the sahelian drought of the 1970s and 1990s also contributed to increased dependence on

groundwater within the region as groundwater became the only reliable source of water

during those periods.

The combined effects of climate variability, the generally low yields of crystalline basement

rocks and heavy dependence on groundwater has often resulted in the drying up of shallow

wells and boreholes, and decline in groundwater levels, which eventually culminates in

higher pumping heads and greater costs of pumping. This situation therefore raises a lot of

concerns about the sustainable development, use and management of groundwater in the

region. As a starting point in the sustainable management of this vital resource, working

estimates of the rates of groundwater recharge is required.

Numerous techniques and models have been developed over the years, to estimate natural

groundwater recharge. These include the empirical method, groundwater resource estimation

method, groundwater balance approach and soil moisture based methods (Kumar,

2000).Other methods as identified by Kumar & Seethapathi (2002), include groundwater

balance method, soil water balance method, zero flux plane method, one-dimensional soil

water flow model, inverse modeling technique and isotope and solute profile techniques.

A number groundwater recharge studies in which empirical methods were used exists in

literature. For example, Oke et al., (2013) compared the estimated annual groundwater

recharge for the Ogun-Oshun River Basin using three empirical methods, namely the Krishna

Rao formula, the Kumar and Seethapathi formula and the modified Chaturvedi formula. The

results of the study showed that the annual recharge estimate based the first two formulae

were similar while the estimates based on the modified Chaturvedi formula differed

significantly.

Utilising the Amristsar formula in combination with remote sensing techniques, Rawat et al.,

(2012) estimated groundwater recharge at Shankergarh block in Allahabad, India. The result

of the study revealed a close correlation between the empirical method employed and the

estimate obtained through remote sensing. Adeleke et al., (2015) used the modified

Chaturvedi formula to estimate groundwater recharge for Odeda Local Government Area of

Ogun State between 1983 and 2014. The result of the study showed that the rate of recharge

was low with 11 percent of the rainfall received percolating into the underlying aquifers.

Furthermore, a number of studies exist on groundwater recharge in Northern Nigeria in which

different estimation methods were used. These include the adoption of the water budget

method in estimating groundwater recharge in the Hadejia-Nguru wetlands (Goes, 1999), the

use of the chloride mass balance method in estimating groundwater recharge in Northeast

Nigeria (Edmunds, et al, 2002), use of the unsaturated zone chloride profile technique in

estimating groundwater recharge in Northeast Nigeria (Goni, 2001) and the utilization of a

combination of empirical, chloride mass balance, climatic-hydrological balance, stream flow

hydrographs and water table recharge methods, in estimating groundwater recharge in the

parts of Sokoto Basin (Adelana, et al, 2006).

In recognition of the high variability of this natural flux and the important role that

groundwater availability plays in the economic sustenance of Northern Nigeria, this study

aims to estimate natural groundwater recharge at Gusau (Northwest Nigeria) and Bauchi


60

(Northeast Nigeria) located in the basement complex of Northern Nigeria using empirical

methods. Secondly, the study aims to determine the level of variability of annual rainfall and

the recharge estimates. Thirdly, the study aims to determine the existence or otherwise of

equality or homogeneity in the variance of the recharge estimates derived from the empirical

methods. Lastly the study aims to assess the extent to which rainfall accounts for variation in

recharge over the study period as well as predict the mean groundwater recharge for the study

period using the simple linear regression method.

MATERIALS AND METHODS

The study locations are Gusau and Bauchi. Gusau (Fig.1) is located between latitudes

12013′N and 12

018′N and between longitudes 6

029′E and 6

045′E (Dalhatu & Garba, 2012).

Gusau Local Government Area in which Gusau Town is located occupies an area of 3, 364

Km2 (Ladan, et al., 2012). Gusau is located in the Sudan Savanna ecological zone of Nigeria.

Climate-wise the dry season lasts from November to April, while the rainy season is from

May to October. The rainfall pattern is characterized by a single peak, with an annual mean

of about 875mm, while mean annual temperature is around 300C (Dalhatu & Garba, 2012).

The hydrographic features of Gusau area include River Sokoto and some of its tributaries, as

well as some of the westward and southward flowing tributaries of Rivers Rima and Niger

(Shear & Partners as cited in Dalhatu & Garba, 2012). Relief-wise Gusau is located at an

altitude of 300m above sea level (Dalhatu & Garba, 2012). The topography of the area is

undulating and slopes gently into the river valley (Swindel as cited in Dalhatu & Garba,

2012).

Bauchi town is located between latitudes 9000′N and 9

030′N and between longitudes 10

025′E

and 11020′E. It occupies a total land area of 3, 604 hectares (36.04Km

2). Bauchi is located in

the Sudan Savanna zone of Nigeria. The landscape is relatively flat, with a range of disjointed

hills at the northern part of the town (Usman & Mohammed, 2012).

Rainfall in Bauchi is strongly seasonal, starting mostly in the first half of May and ending in

September. Rain days vary between 100 and 160, while annual rainfall is about 1,000mm.

Highest temperatures are recorded in April and May prior to the onset of the rainy season

when mean monthly is about 29.70C while mean monthly temperature drops to 22.2

0C in the

month of December (Hill & Wall, 1978; Hazell, 1985).


61

Bauchi and Gusau are underlain by the basement complex rocks of Northern Nigeria. The

basement complex rocks are mostly made up of gneisses, migmatites and granites, with most

of the area underlain by this rock characterized by a thin, discontinuous mantle of weathered

rock (Du Preez & Barber, 1965).

Basement complex rocks are generally poor aquifers. The occurrence of groundwater

depends on the thickness and extent of the decomposed mantle or the presence of joints and

fractures in the fresh rock, with groundwater existing either in the weathered mantle or in the

joints and fracture system in the unweathered rocks.

Study on the groundwater resources development of Bauchi State (Hazell, 1985) showed that

the crystalline basement complex rocks have variable permeability, low water storage

capacity, with the depth to water table in the weathered zone ranging from 3 to 18m, while

the depth ranges between 25 and 40m where there are fractures and joints. According to

Mustafa and Adamu as cited in Dike et al (2008), hydraulic conductivity of basement

complex aquifers in Bauchi Town ranges between 0.09 and 0.46m/day, while transmissivity

is less than 2m2/day. Average borehole depth is about 32m, with a yield of about 0.8m/s

recorded in some parts of the metropolis. With regards to groundwater occurrence at Gusau,

boreholes drilled into the pre-cretaceous crystalline rocks 23 miles from Gusau showed that

the thin weathered veins made up of quartz fragments, sand and clay fillings were found to

contain most of the exploitable water at depths ranging between 94 and 135 ft.

Fig. 1: Locations of Bauchi and Gusau


62

Data used for this study are rainfall and temperature data for the period 2000 to 2010. The

data were obtained from the archives of the Nigerian Meteorological Agency (NIMET). Both

datasets served inputs into the groundwater recharge equations. Two empirical methods were

utilised in the estimation of groundwater recharge for this study. The first method was

developed by Sinha and Sharma (1988). The method utilises rainfall as the only variable and

it is applicable to semi-arid environments with rainfall greater than 380mm/year (Adelana et

al., 2006). The formula as utilised in the works of Adelana et al (2006) is expressed as:

r = 50.8 (p/25.4-15)0.4

………………………………………….(1)

Where

r = Recharge

p = Precipitation

The second empirical method is based on Turc’s (1954) empirical method for estimating

recharge. The formula is expressed as:

r = P (1-[0.9+P2/L

2]

-0.5) ………………………………..(2)

Where

r = Annual average recharge (mm/yr)

p = Annual precipitation (mm/yr)

T = Mean annual temperature (0C)

L= 300+25T+0.05T2 …………………………………………...(3)

The two methods were adopted because empirical methods have been found to be useful in

areas with inadequate hydrogeological data. Secondly, both methods were developed to

estimate groundwater recharge in similar climatic and hydrogeological environments. Both

methods were adopted as used in Adelana et al (2006), due to the similarities in

environmental conditions, as Gusau and Bauchi are located in the same semi-arid region in

which the Sokoto basin is located.

In order to determine the level of variability of annual rainfall and the estimated recharge

over the 11-year study period, the coefficient of variability was used. This is expressed as:

CV (%) �

�*100 ………………………………………….. (4)

where

σ = Standard deviation of the data series

X = Mean of the data series.

Levene’s test was used to test the equality of variance between the recharge estimates derived

from the two empirical methods. This is expressed as:

� = (� = (� − �) n�(Z��Z)�

��2

(g − 1) ∑��

�

��(Z��

− Z�)2 ………………………………… (5)

Where

�� = |�� − ��| ………………………………………………………………… (6)

Zk=�

��∑ ��

�� ………………………………………………………………… (7)

� =�

! ∑ Zki$�

%��

�� …………………………………………………………… (8)

Yk=�

��∑ ��

�� ………………………………………………………………… (9)

Simple linear equation was used to predict the mean groundwater recharge for the study

period. The equation of the simple linear regression is expressed as:


63

Ƴ = (b0 + b1X1 ) + ɛi ............................................................................... (10)

where

Ƴ = Dependent variable predicted by regression model

b0 = Intercept

b1 = ith coefficient of Xi

Xi= ith independent variable from total set of q variables

ɛi = Random errors

RESULTS AND DISCUSSION Recharge Estimates

The annual rainfall amount and the results of recharge estimates based on Sinha and Sharma

(1988) and Turc (1954) for Bauchi and Gusau is as presented in Table 1.

Table 1: Annual Rainfall Amount and Recharge Estimates for Bauchi and Gusau

Bauchi Gusau

Year Annual

Rainfall

(mm)

Recharge Estimates

(mm)

Annual

Rainfall

(mm)

Recharge Estimates

(mm)

Sinha

and Sharma

( Method 1)

Turc

(Method 2)

Sinha

and Sharma

( Method 1)

Turc

(Method 2)

2000 1058.9 188.98 320.95 862.1 208.28 194.58

2001 1300.8 213.36 501.46 732.8 145.29 125.24

2002 819.1 158.50 170.13 889.4 210.82 210.70

2003 989.5 180.85 263.40 1416.1 252.49 578.76

2004 865.5 165.1 188.07 755.4 197.10 129.78

2005 1034.5 186.44 292.97 920.3 172.21 220.23

2006 1017.9 184.40 281.96 961.8 217.42 236.12

2007 1136.9 197.61 376.54 615.8 181.86 69.52

2008 1133.1 197.10 361.23 954.0 216.41 271.70

2009 1531.3 261.62 674.23 1006.0 221.49 284.80

2010 1547.0 262.64 687.49 1035.9 224.03 283.11

For Bauchi the results of the natural groundwater recharge estimates based on method 1

(Sinha & Sharma, 1988) shows that the periods of highest and lowest recharge coincided with

the period of highest and lowest annual rainfall amount. In 2010 when the highest annual

rainfall amount of the series (1547mm) was recorded, the highest recharge estimate of

262.64mm was also computed. For the year 2002 when the lowest annual rainfall amount

(819.1mm) of the series was recorded, the lowest recharge estimate of 158.50mm was also

computed.

The results of the groundwater recharge estimation based on the second method (Turc, 1954)

also shows that periods of highest and lowest annual average recharge estimates coincided

with periods of highest and lowest annual rainfall. Year 2010 in which the highest annual

rainfall of the series was recorded coincided with the year with the highest annual average

groundwater recharge estimate of 687.49mm. Similarly, year 2002 when the lowest annual

rainfall of the series was recorded coincided with the period of the lowest annual average

recharge estimate of 170.13mm.


Furthermore, most of the estimated annual recharge based on Sinha and Sharma method

(Fig.2) and Turc method (Fig.3) were below the 11

respectively.

Fig.2: Annual and Long Term Mean of Recharge Estimate at Bauchi based on Sinha and

Sharma Method

Fig.3: Annual and Long Term Mean of Recharge Estimate at Bauchi based on Turc Method

For Gusau, results of the recharge estimates

highest annual rainfall amount of the series (rainfall amount of 1416.1mm for 2003)

coincided with the period of the highest annual recharge estimate of 252.49mm. On the other

hand however, the period with the lo

2007) did not coincide with period of the lowest annual recharge estimate of 142.29mm

which was computed for 2001. The year 2001 for which the lowest annual recharge estimate

was computed coincided with the period with the second lowest annual rainfall of the series.

High potential evapotranspiration may a contributive factor to the period of lowest annual

rainfall not coinciding with the period for which lowest groundwater recharge was computed.

For the second recharge method, periods of highest and lowest annual rainfall coincided with

periods of highest and lowest annual average recharge estimates. Year 2003 in which the

highest rainfall amount of the series was recorded coincided with the period of

annual average recharge estimate of 578.76mm. The same scenario was established for the

period of lowest annual rainfall and lowest annual average recharge estimate. The period with

of Geographic Issues, Vol. 2 No. 1, 2014

64


(Fig.2) and Turc method (Fig.3) were below the 11-year mean of 199.7 mm and 37



For Gusau, results of the recharge estimates based on method 1 show that the period with



hand however, the period with the lowest annual rainfall of the series (615.8mm recorded in



th the period with the second lowest annual rainfall of the series.



the second recharge method, periods of highest and lowest annual rainfall coincided with


highest rainfall amount of the series was recorded coincided with the period of




year mean of 199.7 mm and 374.4 mm



based on method 1 show that the period with



west annual rainfall of the series (615.8mm recorded in



th the period with the second lowest annual rainfall of the series.



the second recharge method, periods of highest and lowest annual rainfall coincided with


highest rainfall amount of the series was recorded coincided with the period of the highest




the lowest annual rainfall of the series (year 2007) coincide

annual average recharge estimate of 69.52mm.

Comparison of the estimated annual recharge with the long term mean based on Sinhna and

Sharma method shows that 7 out the 11 years had recharge estimates higher than the mean

value of 204.3mm (Fig.4). For Turc method (Fig.5) an almost equal number of years had

annual recharge estimates lower than and higher than the long term mean of 236.8mm.

Fig.4: Annual and Long Term Mean of Recharge Estimate at Gusau based on Sinha and

Sharma Method

Fig.5: Annual and Long Term Mean of Recharge Estimate at Gusau based on Turc Method

Variability and Equality of VarianceThe coefficient of variation for rainfall and recharge estimates for both locations is as

presented in Table 2.

Table 2: Coefficient of Variation of Rainfall and Recharge Estimates

Location Rainfall (mm)

Bauchi 21%

Gusau 22%

of Geographic Issues, Vol. 2 No. 1, 2014

65

the lowest annual rainfall of the series (year 2007) coincided with the period with the lowest

annual average recharge estimate of 69.52mm.



alue of 204.3mm (Fig.4). For Turc method (Fig.5) an almost equal number of years had




Variability and Equality of Variance The coefficient of variation for rainfall and recharge estimates for both locations is as

: Coefficient of Variation of Rainfall and Recharge Estimates

Rainfall (mm) Sinha and Sharma (Method 1) Turc (Method 2)

17 % 47%

14% 56%

d with the period with the lowest



alue of 204.3mm (Fig.4). For Turc method (Fig.5) an almost equal number of years had




The coefficient of variation for rainfall and recharge estimates for both locations is as

Turc (Method 2)


66

As shown in the Table the CV for both locations based on the Turc method is much higher

than coefficient of variation based on Sinha and Sharma method. This is an indication that

CV derived based on Turc method is more variable. Due to the inherent variability of climatic

and surface conditions, temporal variation of groundwater recharge occurs over different

temporal scales. These variations are normal and they depict the dynamic balance between

precipitation, infiltration, evaporation, transpiration, runoff, groundwater storage, withdrawal

and base-flow discharge (Resse and Risser, 2010). Apart from these factors, drought,

irrigation, agriculture and other forms of land use changes may also contribute to the

temporal variation of recharge within the study locations.

The comparison of the CV of rainfall with CV of recharge estimates for both methods shows

a closer relationship between rainfall and recharge estimate based on Sinha and Sharma

method.Results of the Levene’s test for equality of variance for Bauchi and Gusau are

represented Tables 3 and 4 respectively.

Table 3: Levene’s Test for Equality of Variance at Bauchi

Levene

Statistic df1 df2 Sig.

Recharge Based on Mean 11.462 1 20 .003

Based on Median 7.032 1 20 .015

Based on Median and

with adjusted df

7.032 1 10.859 .023

Based on trimmed

mean

10.716 1 20 .004

Table 4: Levene’s Test for Equality of Variance at Gusau

Levene

Statistic df1 df2 Sig.

Recharge Based on Mean 4.484 1 20 .047

Based on Median 4.145 1 20 .055

Based on Median and

with adjusted df

4.145 1 10.824 .067

Based on trimmed

mean

4.240 1 20 .053

For Bauchi, F(1, 20) = 11.46, P <0.05, where P = 0.003 and Gusau, F(1, 20) = 4.48, P <0.05

where P = 0.047, the assumption of equality or homogeneity of variance has been violated ,

indicating that the variance of the recharge estimates obtained from Sinha and Sharma

method, and those obtained from Turc method are significantly different from each other.

Regression Analysis The results of the regression model for Bauchi based on annual rainfall and recharge

estimates computed using the first empirical method is presented in Tables 5a to 5c, while the

regression model obtained from annual rainfall and recharge estimates from the second

empirical method is presented in Tables 6a to 6c.


67

As shown in Table 5a, R2 equals 0.984 indicating that rainfall accounts for 98.4 percent of

variation in natural groundwater recharge for the study period. Other factors that could be

responsible for the variability include vegetation and soil properties. The F-ratio (552.799) in

Table 5b is significant at p < 0.01 because the value in the column labeled sig (last column) is

less than 0.01. This indicates that there is less than 1 percent chance that an F-ratio this large

could be a chance occurrence, leading to the conclusion that the regression model results in

significantly better prediction.

The value of b0 (39.59) in Table 6c indicates that when there is no rainfall (when Xi= 0), the

model predicts that recharge will be 39.59mm while the value of b1 (Table 6c) which

represents the gradient of the regression line indicates that if there is an increase in rainfall by

1mm, then the regression model predicts that recharge will increase by 0.142mm.

The mean rainfall value of 1,130.41mm and the regression equation Ƴ = (b0 + b1X1) + ɛi was

used to predict the mean recharge for the 11-year study period, When:

Y = Mean recharge for the 11-year period

b0 = 39.59mm

bi = 0.142mm

Xi = 1, 130.41mm

is expressed as:

Mean recharge = (39.59 + 0.142 x 1, 130.41) = 200.11mm

Table 5a: Summary of Regression Model

Table 5b: Analysis of Variance, Sums of Squares, F-Ratio and Associated Significance Value

ANOVAb

Model

Sum of

Squares df Mean Square F Sig.

1 Regression 11579.488 1 11579.488 552.799 .000a

Residual 188.523 9 20.947

Total 11768.011 10

a. Predictors: (Constant), Rainfall

b. Dependent Variable: recharge

Table 5c: Beta Values of Regression Model

Coefficientsa

Model

Unstandardized

Coefficients

Standardized

Coefficients

t Sig. B Std. Error Beta

1 (Constant) 39.590 6.948 5.698 .000

Model Summary

Model R R Square

Adjusted R

Square

Std. Error of

the Estimate

1 .992a .984 .982 4.57679



68

Rainfall .142 .006 .992 23.512 .000

a. Dependent Variable: recharge

With regards to the outcome of the regression model of annual rainfall and recharge estimates

obtained from the second empirical method, Table 6a shows that rainfall accounts for 99

percent (R2= 0.996) of the variation in groundwater recharge. Table 6b indicates that the F-

ratio is significant at P < 0.01, while b0 predicts that when there is no rainfall (when Xi = 0),

the recharge value will be -453.70mm. Such negative value is however impossible in nature.

This is because during periods of rainfall cessation such as the dry season, natural

groundwater recharge can occur from seepage from surface water bodies such as streams,

lakes or ponds, irrigation return flow, inter-aquifer flows and urban recharge such as leaky

water and sewage systems. bi which is the slope of the regression line indicates that there will

be an increase of 0.733mm in recharge for every 1mm increase in rainfall.

The prediction of the 11-year mean recharge based on a mean rainfall value of 1,130.41mm

(ie Xi = 1, 130.41mm) and the regression equation Ƴ = (b0 + b1X1) + ɛi , when:


b0 = -453.70mm

bi = 0.733mm

is expressed as:

Mean recharge = (-453.70 + 0.733 x 1, 130.41) = 374.89mm


Model R R Square

Adjusted R

Square

Std. Error of

the Estimate

1 .998a .996 .996 11.80280


Table 6b: Analysis of Variance, Sums of Squares, F-Ratio and Associated Significance

Value

ANOVAb

Model

Sum of


1 Regression 309790.897 1 309790.897 2223.815 .000a

Residual 1253.755 9 139.306

Total 311044.652 10


b. Dependent Variable: Recharge


Coefficientsa

Model

Unstandardized

Coefficients

Standardized

Coefficients



69

1 (Constant) -453.700 17.917 -25.322 .000

Rainfall .733 .016 .998 47.157 .000

a. Dependent Variable: Recharge

The results of the regression model for Gusau based on annual rainfall and recharge estimates

computed using the first empirical method is presented in Tables 7a to 7c, while the

regression model obtained from annual rainfall and recharge estimates computed with the

second empirical method is presented in Tables 8a to 8c.

As shown in Table 7a, the value of R2 (0.616) indicates that rainfall accounts for 61.6 percent

of the variation in natural groundwater recharge. The F-ratio (14.447) as shown in Table 7b is

significant at P < 0.01 because the value 0.004 in the column labeled sig is less than 0.01.

The value b0 (102.612) in Table 8c indicates that when there is no rainfall (when Xi = 0), the

model predicts a recharge of 102.62mm.The slope of the regression line (bi) in Table 7c

indicates that when there an increase of 1mm in rainfall, groundwater recharge will increase

by 0.11mm.

The prediction of the 11-year mean recharge based on a mean rainfall value of 922.70mm (ie

Xi = 922.70) and the regression equation Ƴ = (b0 + b1X1) + ɛi , when:


b0 = 102.62mm

bi = 0.11mm

is expressed as:

Mean recharge = (102.62 + 0.11 x 922.70) = 204.12mm


Table 7b: Analysis of Variance, Sums of Squares, F-Ratio and Associated Significance Value

ANOVAb

Model Sum of Squares df Mean Square F Sig.

1 Regression 5207.472 1 5207.472 14.447 .004a

Residual 3244.051 9 360.450

Total 8451.523 10


Coefficientsa

Model

Unstandardized

Coefficients

Standardized

Coefficients


1 (Constant) 102.619 27.359 3.751 .005

Model Summary

Model R R Square

Adjusted R

Square

Std. Error of

the Estimate

1 .785a .616 .574 18.98552




70

Rainfall .110 .029 .785 3.801 .004


For the regression model of annual rainfall and recharge estimates obtained from the second

empirical method, Table 8a indicates that rainfall accounts for 98 percent (R2 = 0.980) of the

variation in natural groundwater recharge for the study period. The F-ratio (448.948) in Table

8b is significant at P < 0.01 because the value 0.000 under the column labeled sig is lesser

0.01, while the value of b0 predicts that when there is no rainfall (when Xi = 0), the recharge

value will be -350.60mm. The negative value however, does not mean the cessation of

recharge for reasons earlier explained. The value of bi predicts that for every 1mm increase in

rainfall, there will be a corresponding increase in recharge by 0.637mm.

The prediction of the 11-year mean recharge based on a mean rainfall value of 922.70mm (ie

Xi = 922.70) and the regression equation Ƴ = (b0 + b1X1) + ɛi , when:


b0 = -350.597mm

bi = 0.637mm

is expressed as:

Mean recharge = (-350.59 + 0.637 x 922.70) = 237.16mm


Table 8b: Analysis of Variance, Sums of Squares, F-Ratio and Associated Significance

Value


Coefficientsa

Model

Unstandardized

Coefficients

Standardized

Coefficients


1 (Constant) -350.597 28.349 -12.367 .000

Rainfall .637 .030 .990 21.188 .000

Model Summary

Model R R Square

Adjusted R

Square

Std. Error of

the Estimate

1 .990a .980 .978 19.67218


ANOVAb

Model

Sum of


1 Regression 173740.586 1 173740.586 448.948 .000a

Residual 3482.953 9 386.995

Total 177223.540 10


b. Dependent Variable: Recharge


71

Coefficientsa

Model

Unstandardized

Coefficients

Standardized

Coefficients


1 (Constant) -350.597 28.349 -12.367 .000

Rainfall .637 .030 .990 21.188 .000


CONCLUSION AND RECOMMENDATION The study has shown that empirical methods can be used as a plausible means of estimating

groundwater recharge. The moderate (Sinha and Sharma Method) to high level (Turc

Method) of variability as shown by the coefficient of variation is an indication that the

variable nature of recharge should be given more consideration in the development of

groundwater. This would help in preventing groundwater mining and ensure sustainable

water use.

The study also showed through regression analysis that rainfall generally accounts for more

than 90 percent of groundwater variation in the study locations. This therefore shows that

statistical methods such as regression analysis can serve as a veritable means of predicting the

rate of groundwater recharge.

Going by the results of this study, it is recommended that scientific and periodic assessments

of groundwater resources potential should be embarked upon. Furthermore, more efforts

should be aimed at water conservation through rain water harvesting in the wet season to

reduce the level of reliance on groundwater.

REFERENCES

Adelana, S.M.A.., Olasehinde, P.I. and Vrbka, P. (2006). A quantitative estimation of

groundwater

recharge in part of the Sokoto Basin, Nigeria. Journal of Environmental Hydrology,

14(5), 1-16.

Adeleke, O.O., Makinde, V., Eruola, A.O., Dada, O.F., Akintayo, O.O., and Aluko, T.J.

(2015).

Estimation of groundwater recharges using empirical formulae in Odeda Local

Government Area, Ogun State Nigeria. Challenges, 6, 271-281

Dalhatu, S., and Garba, J. (2012). Soil resources degradation and conservation techniques

adopted

among the small holder farmers in Gusau, North-western Nigeria. Nigerian Journal of

Basic and Applied Sciences, 20 (3), 134-141.

De Vries., J.J. and Simmers, I. (2002). Groundwater recharge: An overview of process and

challenges. Hydrogeology Journal, 10, 5-17.

Dike, E.F.C., Aga, T. and Bashayi, T.E.O. (2008). Geochemical investigation of groundwater

in

shallow hand dug wells in wells in railway area, Bauchi. Continental Journal of

Applied Sciences, 3, 57-64.


72

Du Preez, J.W., and Barber, W. (1965). The distribution and chemical quality of

groundwater in

Northern Nigeria. (Lagos: Federal Government of Nigeria).

Edmunds, W.M.., Fellman, E. Goni, L.B. and Prudhomme, C. (2002). Spatial and temporal

distribution of groundwater recharge in northern Nigeria. Hydrogeology Journal, 10,

205-215.

Goes, B.J.M., (1999). Estimate of shallow groundwater recharge in the Hadejia-Nguru

wetlands,

Semi-arid, Northeast, Nigeria. Hydrogeology Journal, 7, 305-316.

Hazell, J.R.T., (1985). Development of groundwater for farming communities in Bauchi

State,

Nigeria. In hydrogeology in the service of man. Memoirs of the 18th

congress of the

International Association of Hydrogeologists, 8-13 September, 1985, Cambridge:

IAHS Publication No 154.

Hill, I.D. (Ed). (1978). Land resources of central Nigeria: Agricultural development

possibilities. Vol 1B. The Bauchi Plains. Surrey: Land Resources Development

Centre.

Kumar, C.P. (2000). Groundwater assessment methodology. National Institute of Hydrology.

Kumar, C.P., and Seethapathi, P.V. (2002). Assessment of natural groundwater recharge in

upper

Ganga command area. Journal of Applied Hydrology, 14(4), 13-20.

Ladan, M.U., Abubakar, U. Abdullahi, K. Bunza. M.D.A. Ladan, M.J. and Adamu, T.

(2012).

Studies on urinary schistosomiasis in selected villages around Gusau Dam site,

Zamfara State. Nigeria. Nigerian Journal of Basic and Applied Sciences, 20(3), 189-

194.

Oke, M.O., Martins, O., Idowu, O., and Aiyelokun, O. (2013). Comparative analysis of

empirical

formulae used in groundwater recharge in Ogun-Oshun River Basin. Journal of

Scientific Research and Reports, 2(2), 692-710.

Rawat, K.S., Mishra, A.K., Paul, G., Kumar, R. (2012). Estimation of groundwater recharge

in Shankergarh block of Allahabad, India using remote sensing and statistical approach.

Global Journal of Science, Engineering and Technology, 1, 34-48.

Sharp, J.M., (2007). The glossary of hydrogeological terms. The University of Texas, Texas.

Usman, H.A., and Mohammed, B.U. (2012). Solid waste management in Bauchi, Nigeria

(Obstacles and prospects). Journal of Environmental Science and Resources

Management, 4, 96-107.

Documents

EMPIRICAL ESTIMATION OF GROUNDWATER RECHARGE IN … · Olubunmi Adegun Department of Geography, University of Lagos, Lagos, Nigeria Correspondence:[email protected] ... Nigeria (Edmunds,